Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →YAML front matter (YFM) for any markdown article. x-cmd's default is three top-level fields: `tags`, `description`, `memo`. Default borrows the Agent Skills frontmatter shape — see <https://agentskills.io/specification>. The yfm module is a helper: `x yfm ls / init / lint`.
| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-01 | ✗→✓ | ▲ Improved | -1% | 0% |
| case-02 | ✗→✓ | ▲ Improved | -64% | 0% |
| case-03 | ✗→✓ | ▲ Improved | 29% | 0% |
| case-04 | ✗→✓ | ▲ Improved | -47% | 0% |
| case-05 | ✗→✓ | ▲ Improved | -49% | 0% |
Use YFM for any markdown an agent, index, or ontology pipeline may scan. Follow the existing dialect if one is already in use (Agent Skills, Obsidian, Jekyll, Hugo, etc.) — the x-cmd default applies only when no other dialect governs.
yaml--- tags: [design, front-matter] description: One-sentence summary a reader uses to decide whether to open this file. memo: 2026-07-31T14:30:00+08:00: Cut aliases; renamed third field to memo. --- # Title Body...
tags — top-level list of slugs [a-z0-9-]. Consumer: the collector for ontology indexing.description — top-level string. One sentence. Not the Agent Skills description: (that is a SKILL.md-only requirement, governed by skill0-writer).memo: — top-level map, the file's revision log. Keys are ISO 8601 timestamps (YYYY-MM-DDTHH:mm:ss±HH:mm, author's local offset); values are short change-descriptions. Newest first. Omit when no history is worth recording.<project>/.x-cmd/story/ — extends with issue: — see usecase/story.md.repo:, number: — see usecase/git-issue-pr-wiki-markdown-etc.md.README.md — default fields only.The yfm module is a helper. Three commands:
x yfm ls [path] — list YFM blocks; defaults to the current directoryx yfm init [path] — write an x-cmd-default block into each markdown file under the pathx yfm lint [path] — check that each YFM block is valid YAML and follows the dialect it claimsThe collector ingests many dialects. For the spec the default borrows from, see <https://agentskills.io/specification>. Per-dialect:
metadata: is string→string; top level is closedtags / aliases / cssclasses reserved| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→pass | 8,321 | 4,766 | -43% | 1 | 1 | 0% | 1,376 | 1,359 | -1% | 0 | 0 | — |
case-02 | fail→pass | 19,065 | 3,046 | -84% | 1 | 1 | 0% | 2,876 | 1,032 | -64% | 0 | 0 | — |
case-03 | fail→pass | 14,290 | 5,859 | -59% | 1 | 1 | 0% | 1,136 | 1,461 | +29% | 0 | 0 | — |
case-04 | fail→pass | 11,794 | 3,410 | -71% | 1 | 1 | 0% | 1,922 | 1,025 | -47% | 0 | 0 | — |
case-05 | fail→pass | 42,117 | 2,868 | -93% | 1 | 1 | 0% | 2,034 | 1,038 | -49% | 0 | 0 | — |
case-06 | fail→fail | 15,512 | 7,590 | -51% | 1 | 1 | 0% | 996 | 1,699 | +71% | 0 | 0 | — |
case-07 | fail→pass | 21,636 | 5,605 | -74% | 1 | 1 | 0% | 1,059 | 1,517 | +43% | 0 | 0 | — |
case-08 | fail→pass | 6,256 | 5,327 | -15% | 1 | 1 | 0% | 898 | 1,473 | +64% | 0 | 0 | — |
case-09 | fail→pass | 11,917 | 5,507 | -54% | 1 | 1 | 0% | 1,925 | 1,666 | -13% | 0 | 0 | — |
case-10 | fail→pass | 6,703 | 4,874 | -27% | 1 | 1 | 0% | 1,226 | 1,592 | +30% | 0 | 0 | — |
case-11 | fail→pass | 21,118 | 7,406 | -65% | 1 | 1 | 0% | 1,757 | 1,791 | +2% | 0 | 0 | — |
case-12 | pass→pass | 10,725 | 3,812 | -64% | 1 | 1 | 0% | 1,561 | 1,190 | -24% | 0 | 0 | — |
case-13 | pass→pass | 21,350 | 6,046 | -72% | 1 | 1 | 0% | 1,930 | 1,410 | -27% | 0 | 0 | — |
case-14 | pass→pass | 4,484 | 3,967 | -12% | 1 | 1 | 0% | 637 | 1,197 | +88% | 0 | 0 | — |
case-15 | pass→pass | 7,090 | 8,366 | +18% | 1 | 1 | 0% | 1,176 | 1,196 | +2% | 0 | 0 | — |
case-16 | fail→pass | 7,422 | 3,949 | -47% | 1 | 1 | 0% | 1,126 | 1,029 | -9% | 0 | 0 | — |
case-17 | fail→pass | 6,445 | 5,420 | -16% | 1 | 1 | 0% | 950 | 1,363 | +43% | 0 | 0 | — |
case-18 | pass→pass | 21,548 | 13,089 | -39% | 1 | 1 | 0% | 2,400 | 1,830 | -24% | 0 | 0 | — |
case-19 | fail→pass | 39,044 | 10,650 | -73% | 1 | 1 | 0% | 1,409 | 2,229 | +58% | 0 | 0 | — |
case-20 | pass→pass | 10,304 | 9,340 | -9% | 1 | 1 | 0% | 1,930 | 2,247 | +16% | 0 | 0 | — |
case-21 | pass→pass | 7,919 | 7,313 | -8% | 1 | 1 | 0% | 1,235 | 1,888 | +53% | 0 | 0 | — |
case-22 | pass→pass | 11,265 | 10,452 | -7% | 1 | 1 | 0% | 2,008 | 2,248 | +12% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 22 cases were attempted, and 20 counted toward the lift figure. The other 2 produced results that are not comparable between the two arms, so they are excluded from the headline rather than averaged into it. The headline lift of +59 percentage points is the difference between those two pass rates over the 20 comparable cases.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.